CUDA GPU Kernel Performance Engineer (Contract)

Mercor; United States; Remote; USD 96,432 - 179,088; Part time

Job summary

Mercor is seeking GPU kernel optimization experts to contribute to a cutting-edge project with a leading AI lab. This contract-based role targets freelancers with strong C++17 skills, practical GPU programming experience, and the ability to improve kernel performance through profiler-guided analysis.

You will evaluate, optimize, and reason about GPU kernels across modern hardware environments, writing and refining C++17, Python, and GPU code with emphasis on measurable performance gains.

Qualifications

• At least 1 year of GPU experience.
• Strong C++17 and Python knowledge.
• Experience with CUDA, HIP or similar GPU models.
• Able to work 20 hrs/week as a contractor.
• Familiarity with GPU profiler metrics for optimization.

Responsibilities

• Analyze and optimize GPU kernels for performance and hardware utilization.
• Use profiler metrics to guide kernel improvements.
• Review kernel implementations and identify bottlenecks without deep algorithm context.
• Write, modify, and reason about C++17, Python, and GPU programming code.
• Document optimization decisions clearly, including profiler metric usefulness.

Skills

C++17

Python

GPU programming

Profiler tools

CUDA

HIP

Shader programming

Kernel optimization

Tools

Nsight Compute

CUDA C++ Libraries

Inline PTX

HIP toolkit

Git

Job description

Mercor is seeking GPU kernel optimization experts to contribute to a cutting-edge project with a leading AI lab. This contract-based role targets freelancers with strong C++17 skills, practical GPU programming experience, and the ability to improve kernel performance through profiler-guided analysis.

You will evaluate, optimize, and reason about GPU kernels across modern hardware environments, writing and refining C++17, Python, and GPU code with emphasis on measurable performance gains.